arXiv 2508.031182025-08-05cs.CV

H3R：面向泛化三维重建的混合多视角对应方法

H3R: Hybrid Multi-view Correspondence for Generalizable 3D Reconstruction

ReLER Lab, CCAI, Zhejiang University（浙江大学ReLER实验室，中国计算机协会，浙江大学）；The State Key Lab of Brain-Machine Intelligence, Zhejiang University（浙江大学脑机智能国家重点实验室）；Singapore University of Technology and Design（新加坡科技设计大学）

Heng Jia, Linchao Zhu, Na Zhao

备注：: ICCV 2025

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出H3R混合框架，结合体积潜在融合与注意力聚合，解决多视角对应权衡，收敛快2倍，在多个基准上取得SOTA性能。

摘要：

尽管前馈三维高斯泼溅技术近期取得了进展，泛化三维重建仍具挑战性，尤其是在多视角对应建模方面。现有方法面临一个基本权衡：显式方法实现了几何精度，但在模糊区域表现不佳；而隐式方法提供了鲁棒性，但收敛速度较慢。我们提出H3R，一种混合框架，通过将体积潜在融合与基于注意力的特征聚合相结合来解决这一局限。我们的框架由两个互补组件构成：一个高效的潜在体积，通过极线约束强制几何一致性；以及一个相机感知Transformer，利用Plücker坐标进行自适应对应细化。通过整合这两种范式，我们的方法增强了泛化能力，同时收敛速度比现有方法快2倍。此外，我们表明空间对齐的基础模型（如SD-VAE）显著优于语义对齐的模型（如DINOv2），解决了语义表示与空间重建需求之间的不匹配。我们的方法支持可变数量和高分辨率的输入视图，同时展示了鲁棒的跨数据集泛化能力。大量实验表明，我们的方法在多个基准上达到了最先进的性能，在RealEstate10K、ACID和DTU数据集上分别显著提升了PSNR 0.59 dB、1.06 dB和0.22 dB。代码可在https://github.com/JiaHeng-DLUT/H3R获取。

英文摘要：

Despite recent advances in feed-forward 3D Gaussian Splatting, generalizable 3D reconstruction remains challenging, particularly in multi-view correspondence modeling. Existing approaches face a fundamental trade-off: explicit methods achieve geometric precision but struggle with ambiguous regions, while implicit methods provide robustness but suffer from slow convergence. We present H3R, a hybrid framework that addresses this limitation by integrating volumetric latent fusion with attention-based feature aggregation. Our framework consists of two complementary components: an efficient latent volume that enforces geometric consistency through epipolar constraints, and a camera-aware Transformer that leverages Plücker coordinates for adaptive correspondence refinement. By integrating both paradigms, our approach enhances generalization while converging 2$\times$ faster than existing methods. Furthermore, we show that spatial-aligned foundation models (e.g., SD-VAE) substantially outperform semantic-aligned models (e.g., DINOv2), resolving the mismatch between semantic representations and spatial reconstruction requirements. Our method supports variable-number and high-resolution input views while demonstrating robust cross-dataset generalization. Extensive experiments show that our method achieves state-of-the-art performance across multiple benchmarks, with significant PSNR improvements of 0.59 dB, 1.06 dB, and 0.22 dB on the RealEstate10K, ACID, and DTU datasets, respectively. Code is available at https://github.com/JiaHeng-DLUT/H3R.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

AI总结：

摘要：

英文摘要：

相关论文